A passage in Al Gore’s new book, pointed out to me by Julio Ottino, caught my eye and got me thinking.

In discussing the automation of work, the former Vice President* writes:

And robosourcing is beginning to have an impact on journalism. Narrative Science, a robot reporting company founded by two directors of Northwestern University’s Intelligent Information Laboratory, is now producing articles for newspapers and magazines with algorithms that analyze statistical data from sporting events, financial reports, and government studies. One of the cofounders, Kristian Hammond, who is also a professor at the Medill School of Journalism, told me that the business is expanding rapidly into many new fields of journalism. The CEO, Stuart Frankel, said the few human writers who work for the company have become “meta-journalists” who design the templates, frames, and angles into which the algorithm inserts the data.

Are we all destined to become meta-journalists, meta-physicians, and meta-teachers? And is this a good thing, a bad thing, or more likely, just a thing?

(*Disclosure: I worked for him for a few years in the 1990s, but had no involvement with this book.)

18 Responses to “Anything you can do, I can do meta.”

  1. Rick Carter says:

    Robosourcing will still require human oversight and human wisdom.
    For example, I could (with a little venture capital) create a program that automates litigation (or most of it) for certain types. However, the overall strategy and the precise analysis will require a lawyer. And the client.
    But, as automation does become more pervasive in what had been thought to be exclusively human areas, humans can retrain to became meta-operators and meta-designers.
    The driving force will be that companies will perceive it as more profitable in the short-term and won’t analyze properly the costs of the inevitable massive failure. Because the best system will still be limited by human consciousness. We haven’t yet come up with a self-analyzing system that will identity its weak points.

  2. “We know now that the source of wealth is something specifically human: knowledge.” — Peter Drucker

    Moving up the “food chain” to orchestrate algorithmic processing, or to integrate disciplines into meta-disciplines, affords the robot-displaced former knowledge workers multiple options.

    Innovation affords many more avenues of profitable redeployment.

    Knowledge will remain the source of wealth, but with ever-shorter half lives, constantly renewing human advantages.

    Creating new knowledge that proves itself in action — i.e. by advantageously doing things people want done — will always be in high demand.

  3. What used to be artisanal, one-off solutions, (articles written, surgeries performed, lessons taught) will become the province of experts who have learned how to transfer their highly-developed mental schema over to software for scaled-up delivery. When this happend to musical performance nearly a hundred years ago, recorded music all but snuffed out vernacular (home-grown) music-making. Thus, the best journalists, surgeons, and teachers will become the rock stars of a new age.

  4. I’m all for the automation of the filtering process is concerned when it comes to information gathering. There is far too much information out there to not automate at least some of this process. Knowledge creation requires a human mind to make connections based on experiences. The final part of the journalism equation is the writing itself and I’ve had enough talks with Siri to know that the human element in that conversation is sadly lacking. What makes people want to read is the personality of the author. Writing is not simply about information transmission.

  5. Al Chase says:

    I trust that as we move into the age of meta-journalists and meta-physicians, we will not ignore our need for corporate and civic philosophers to guide our thinking along the right paths. More than ever, we will need to listen to metaphysicians!

  6. If you are the person to whom people look to for guidance, inspiration, knowledge, and insight, and they find out that a machine has been writing your content, good luck with your credibility. (Group “you,” Dan, not you in particular!)

  7. Rajkamal says:

    While automation can churn, analyse, slice and dice data into all sorts of reports – how will it replace the insight, personality, wit, intution that a good “human” writer.

    Maybe this is a thing that will come to be… but it will only push humans into more creative, intuitive “right brain” pursuits.

  8. Xanthe says:

    There’s a similarity here to systems like Nike ID. Instead of pro designers designing shoes, they design a system with which prosumers can “design” their own shoes. So if the pros design meta systems that harness creative capacity of the group formally known as audience, this can be a good thing.

  9. Ganesh says:

    Sorry to point this out, Dan, but you are usually perfect in crediting sources…

    When I saw this title on your tweet, I somehow recalled this being the quoted title statement from legendary software developer Charles Simonyi in an interview with MIT Review many years ago. So I clicked through expecting some reference or connection to that.

  10. Anthony Dina says:

    Um…has anyone read a robot written article? I know that the last time Snap Dragon transcribed some audio text I realized a whole new literary form could be invented. Not to say that we cannot program perfect grammar. Or recount facts. But there’s a whole lot more to the human condition than replay.

  11. John Luchau says:

    Hi Dan

    Eli Pariser actually wrote about this phenomenon in his (frightening..) book “The Filer Bubble”. I would think that Al Gore picked it up there.

    If you have’nt read The Filter Bubble, you better get started… 😉

    (Thanks for the many relevant points you make on your blog!)

  12. I do not consider “robosourcing” a bad thing. In fact, I believe that many human journalists just do that: take data that can be neatly summarised in a visual and dilute it into text that takes far longer to digest and often provides an incomplete picture.

  13. This seems to be contributing to the continued deterioration of what it means to be an expert. We have seen this in manual labor jobs for some time–plumbers, welders, etc. These machines or algorithms, in contrast, are striving to replace intellectual expertise. And if they can, does that redefine the notion of expertise?

    Maybe the prevalence of robots will aid in elevating the meaning of expert in some areas while still degrading others. Could a robot produce one of your books just by aggregating from your sources? Doubtful. But if it could, wouldn’t you just use it as a tool to build an even better book?

  14. Thanks for the thought-provoking post, Dan. I think it’s time for designers to put their minds to creating products and systems that specifically take advantage of human-only capabilities – i.e. their senses and intuition, rather than designing things that are brain-dead simple for humans to use. That way our unique strengths will be utilized and not marginalized.

  15. Ross Arntson says:

    As we take the human element out of businesses and service, our standards of quality will begin to diminish as well.

    This in turn creates an interesting opportunity for many smaller size companies. After-all we all know the frustration we feel when we find ourselves in the infamous telephone loop. BUT frustrated or not, we tolerate it, thus lowering our standard.

Leave a Reply to Dan Pink